Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 최정욱 | - |
dc.date.accessioned | 2022-08-12T04:37:01Z | - |
dc.date.available | 2022-08-12T04:37:01Z | - |
dc.date.issued | 2020-11 | - |
dc.identifier.citation | 2020년도 대한전자공학회 추계학술대회 논문집, Page. 601-604 | en_US |
dc.identifier.uri | https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE10521871 | - |
dc.identifier.uri | https://repository.hanyang.ac.kr/handle/20.500.11754/172399 | - |
dc.description.abstract | Deep Learning Model Quantization is the most effective technique to make a model much lighter and cost efficient in terms of computation. Above many quantization algorithms, PROFIT[1] is a specialized algorithm for sub 4-bit mobile network quantization. But this method has sudden accuracy degradation in 2-bit width precision. In this paper, we propose a better training method to deal with this problem in 2-bit weight quantization. We adopt AIWQ, a metric for the activation’s instability induced by weight quantization [1] and make threshold value with this metric. Using threshold value, we stop training some quantized layers which have high sensitivity to weight quantization and fine-tune the rest of the quantized layers with different learning rate and scheduler. With this advanced training method, we improved 2-bit weight quantization accuracy of light deep learning models including EfficientNetB0 and MobilenetV2. | en_US |
dc.description.sponsorship | 이 논문은 2020년 정부(과학기술정보통신부)의 재원으로 정보통신기획평가원의 지원(No.2020-0-01297, 데이터 재사용 고도화 초저전력 엣지용 딥러닝 프로세서 기술개발), 한국연구재단 – 시스템반도체융합전문인력육성사업의 지원(No. 2020M3H2A107686)과 과학기술정 보통신부 및 정보통신산업진흥원의 ‘고성능 컴퓨팅 지원’ 사업으로부터 지원받아 수행하였음. | en_US |
dc.language.iso | ko_KR | en_US |
dc.publisher | 대한전자공학회 | en_US |
dc.title | 경량 딥러닝 모델의 초저정밀도 양자화를 위한 학습 방식의 개선 | en_US |
dc.title.alternative | Improving training method for very low bit weight quantization of Light Deep Learning Model | en_US |
dc.type | Article | en_US |
dc.relation.page | 601-604 | - |
dc.contributor.googleauthor | 김, 현승 | - |
dc.contributor.googleauthor | 김, 민수 | - |
dc.contributor.googleauthor | 최, 정욱 | - |
dc.sector.campus | S | - |
dc.sector.daehak | COLLEGE OF ENGINEERING[S] | - |
dc.sector.department | SCHOOL OF ELECTRONIC ENGINEERING | - |
dc.identifier.pid | choij | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.